One of the goals of this project was to produce an influenza A H2N2 donor virus containing a PB2 gene possessing attenuating mutations that could be transferred to new epidemic influenza A H1N1 and H3N2 viruses as they emerged in nature so that an antigenically updated live attenuated influenza A virus vaccine could be produced in a timely manner. We applied a new version of reverse genetics developed in our laboratory to the construction of a temperature sensitive (ts) mutant of an influenza A virus that could serve as a donor strain for the purpose just described. This mutant was prepared by introducing two ts mutations sequentially into the PB2 gene that encodes a component of the viral polymerase. One mutation was created at amino acid residue 265 and the other at amino acid residue 658. Although the ts phenotype, initially, appeared to be very stable, we did detect occasional loss of this phenotype in a strict assay for genetic stability in which the mutant was allowed to replicate for 14 days in nude mice. Thus, 2 of 89 isolates obtained from the nasal turbinates of infected nude mice exhibited the ts+ phenotype. The basis for this infrequent event was identified putatively as the development of an extragenic suppressor mutation. In an attempt to create a more stable donor virus, we produced what proved to be our most temperature sensitive mutant (mt) PB2 gene. A separate ts mutation was introduced at amino acid residue 112, 265, or 556. This triple mutation PB2 gene was originally introduced into an influenza A H3N2 virus by transfection of tissue culture cells infected with an H3N2 helper virus and subsequent passage under highly selective conditions that favored the recovery of a reassortant virus bearing the mutant gene. This PB2 gene containing three separate ts mutations was then transferred by gene reassortment into an influenza A H2N2 virus, producing a reassortant which exhibited a high degree of growth restriction in vivo. This reassortant was less ts than the parental triple mt ts PB2 H3N2 transfectant, but it was still highly attenuated in the upper and lower respiratory tract of the hamster. Virus recovered from the hamster was less ts than the parent input virus. The implications of these findings for the future of our influenza virus vaccine development program are discussed.